Businesses of enterprises are rapidly expanding with explosive data growth and the emergence of diverse environments and platforms. As new business environments emerge, more efficient and flexible data services and information processing, and data management functions are needed. In response to these changes, a research is continued on databases to solve problems of high performance, high availability, a high failure recovery capability, and scalability which are foundations of corporate business implementation.
In general, the enterprises use a complicated computing environment constituted by various components including a server, a router, a cloud, a main frame, an application, and a personal computer (PC). In such an environment, an IT system used by the enterprises is configured by various combinations of multiple entities. The respective entities in the IT system may be dependent on each other. Herein, the “IT system” may mean a system for providing various IT services to a user. Further, the “entity” as the component constituting the IT system may include various levels of elements including the application in a unit system. In addition, the term “dependent on each other” means that performance of one entity may influence the performance of another entity.
Since the respective entities of the IT system operate dependently on each other, a fault event which occurs in one entity may concatenatively cause another fault in another entity. Therefore, when the fault occurs in a specific entity, there is an entity that provides a root cause that causes the fault. A technique that determines a root cause entity is a “root cause analysis (RAC)” technique.
In the meantime, various attempts have been made to develop the root cause analysis technique. One among them is an invention regarding a design method of causality model. The related art (Korean Patent Unexamined Application 10-2009-0018806) presents a causality model design method which may express causality of the entities of a system without the need for knowing an overall configuration of the system.
Another related art (Korean Patent Unexamined Application 10-2010-0133168) shows a system and a method that may define the causality which may be used for the root cause analysis from a correlation between application level entities.
The related art has the following disadvantages.
First, a lot of time and efforts of the user are required in defining the causality. In the related art, only when which causality (that is, the correlation including the causality) lower functions which may be included in the application have is defined by the user, causality information between the entities is usable in analyzing the fault root cause. A method that sets the causality defined by the user may increase a trouble of initial system design of the user.
Second, since the related art follows the user defined causality setting method, there is a great dependency on a technical knowledge which the user empirically learns. That is, the fact that the great dependency on the user's knowledge means that accuracy of the causality may vary depending on a skill of the technique and a depth of experience of the user. Accordingly, in this case, since a result for the causality analysis may vary depending on a level of the user, there is a large disadvantage in terms of stability of the system.
Third, the related art cannot flexibly cope with a change in configuration information of the system. When the configuration information of the system is changed, a mutual dependency relation between the entities may Be changed. When the mutual dependency relation is changed, since the user needs to directly set the causality between the entities again, there is a disadvantage in that managing is not easy.
Accordingly, there is a need in the art for tracking and detecting a fault cause of the IT system by a more efficient method.